The good, the bad and the ugly of share of search - A Case Study of Volkswagen

11SCI-VIRUS-TRACKER1-mediumSquareAt3X-12.jpg

“Conventional tracking is slow, expensive, and doesn’t always predict sales. So we’ve been looking at an alternative approach based on search data. It’s fast, cheap & predicts surprisingly far ahead (six months in the example shown). Not only that, it can measure both short and long term ad effects.” - Les Binet

Share of search is the most promising metric of measuring effectiveness. With that being said, with all “new and improved” metrics it’s important to step back and question: what is the metric really saying and what shortcomings are there?

Share of search shows us an output of your marketing initiatives in the form of a behavioral change which is people searching your brand. They do this because:

  1. They want to buy from you

  2. They are already a customer

  3. They are just interested in you

However, point c. is where things can go wrong as pointed out in the case study below: 

Search interest in Volkswagen.

Skärmavbild 2020-10-02 kl. 14.47.08.png

As the graph shows, search for Volkswagen diesel cars went through the roof and was then back to normal after just a couple of weeks.With a peak in search some would say the result should have been a huge sell out, somewhere in September 2015. But as you are probably guessing, this wasn’t a very positive interest! Volkswagen had a massive emission scandal and had a massive drop in sales. 

We argue that share of search as a lone metric to measure brand health is not sufficient. What good is having high search numbers if people’s searches are based on the wrong reasons? If people have a skeptical or negative intention when executing their search it will have a negative impact on both brand and sales.

To remove misinterpretation we need to know the emotions behind the searches: is it a positively or negatively driven interest? Therefore us at My Telescope created the Search & Sentiment metric to tackle this issue: 

Search interest + Sentiment in Volkswagen.

When overlaying the search data with sentiment data, it tells a slightly different story… of a sudden decline and a  somewhat longer recovery.
With My Telescope’s Search & Sentiment index you can see that due to the negative sentiment during this period, there is a clear negative effect on Volkswagen’s brand as well as that this effect took a year to recover from. 

A big thank you to Les Binet for collaborating with us to help us improve this metric.

More about the Search & Sentiment Metric: 

Our “Search & Sentiment index” takes search and amplifies it in the right amount with the ruling sentiment towards the brand. This sentiment is taken from various sources such as news media, social media, forums, blogs and product review sites. The AI driven sentiment analysis, machine learning driven queries and calculations then show us whether people have a genuine interest in the brand and could thus consider to buy. Our Search & Sentiment index hence shows the evolution of the consideration, not the actual numbers.